rename e_above_mp_hull->e_above_hull_mp

Author: janosh

mb_discovery/plot_scripts/hist_classified_stable_as_func_of_hull_dist.py

@@ -36,7 +36,7 @@
     f"{ROOT}/data/2022-06-11-from-rhys/wbm-e-above-mp-hull.csv"
 ).set_index("material_id")
 
-df["e_above_mp_hull"] = df_hull.e_above_mp_hull
+df["e_above_hull_mp"] = df_hull.e_above_hull_mp
 
 
 # %%
@@ -62,7 +62,7 @@
 
 ax = hist_classified_stable_as_func_of_hull_dist(
     e_above_hull_pred=df[pred_cols].mean(axis=1) - df[target_col],
-    e_above_hull_true=df.e_above_mp_hull,
+    e_above_hull_true=df.e_above_hull_mp,
     which_energy=which_energy,
     stability_crit=stability_crit,
     std_pred=std_total,

mb_discovery/plot_scripts/hist_classified_stable_as_func_of_hull_dist_batches.py

@@ -79,7 +79,7 @@
 model_name = "m3gnet"
 df = dfs[model_name]
 
-df["e_above_mp_hull"] = df_hull.e_above_mp_hull
+df["e_above_hull_mp"] = df_hull.e_above_hull_mp
 df["e_form_per_atom"] = df_wbm.e_form_per_atom
 
 
@@ -89,7 +89,7 @@
 
     hist_classified_stable_as_func_of_hull_dist(
         e_above_hull_pred=batch_df.e_form_per_atom_pred - batch_df.e_form_per_atom,
-        e_above_hull_true=batch_df.e_above_mp_hull,
+        e_above_hull_true=batch_df.e_above_hull_mp,
         which_energy=which_energy,
         stability_crit=stability_crit,
         ax=ax,
@@ -101,7 +101,7 @@
 
 hist_classified_stable_as_func_of_hull_dist(
     e_above_hull_pred=df.e_form_per_atom_pred - df.e_form_per_atom,
-    e_above_hull_true=df.e_above_mp_hull,
+    e_above_hull_true=df.e_above_hull_mp,
     which_energy=which_energy,
     stability_crit=stability_crit,
     ax=axs.flat[-1],

mb_discovery/plot_scripts/precision_recall_vs_calc_count.py

@@ -85,14 +85,14 @@
     except AttributeError as exc:
         raise KeyError(f"{model_name = }") from exc
 
-    df["e_above_mp_hull"] = df_hull.e_above_mp_hull
+    df["e_above_hull_mp"] = df_hull.e_above_hull_mp
     df["e_form_per_atom"] = df_wbm.e_form_per_atom
     df["e_above_hull_pred"] = model_preds - df.e_form_per_atom
     if n_nans := df.isna().values.sum() > 0:
         assert n_nans < 10, f"{model_name=} has {n_nans=}"
         df = df.dropna()
 
-    F1 = f1_score(df.e_above_mp_hull < 0, df.e_above_hull_pred < 0)
+    F1 = f1_score(df.e_above_hull_mp < 0, df.e_above_hull_pred < 0)
     F1s[model_name] = F1
 
 
@@ -101,8 +101,8 @@
     df = dfs[model_name]
 
     ax = precision_recall_vs_calc_count(
-        e_above_hull_error=df.e_above_hull_pred + df.e_above_mp_hull,
-        e_above_hull_true=df.e_above_mp_hull,
+        e_above_hull_error=df.e_above_hull_pred + df.e_above_hull_mp,
+        e_above_hull_true=df.e_above_hull_mp,
         color=color,
         label=f"{model_name} {F1=:.2}",
         intersect_lines="recall_xy",  # or "precision_xy", None, 'all'
@@ -113,7 +113,7 @@
 # optimal recall line finds all stable materials without any false positives
 # can be included to confirm all models start out of with near optimal recall
 # and to see how much each model overshoots total n_stable
-n_below_hull = sum(df_hull.e_above_mp_hull < 0)
+n_below_hull = sum(df_hull.e_above_hull_mp < 0)
 ax.plot(
     [0, n_below_hull],
     [0, 100],

mb_discovery/plot_scripts/rolling_mae_vs_hull_dist.py

@@ -39,7 +39,7 @@
     f"{ROOT}/data/2022-06-11-from-rhys/wbm-e-above-mp-hull.csv"
 ).set_index("material_id")
 
-df["e_above_mp_hull"] = df_hull.e_above_mp_hull
+df["e_above_hull_mp"] = df_hull.e_above_hull_mp
 
 assert all(n_nans := df.isna().sum() == 0), f"Found {n_nans} NaNs"
 
@@ -59,7 +59,7 @@
 # %%
 ax = rolling_mae_vs_hull_dist(
     e_above_hull_pred=df.e_above_hull_pred,
-    e_above_hull_true=df.e_above_mp_hull,
+    e_above_hull_true=df.e_above_hull_mp,
     label=legend_label,
 )
 

mb_discovery/plot_scripts/rolling_mae_vs_hull_dist_wbm_batches.py

@@ -28,8 +28,8 @@
     f"{ROOT}/data/2022-06-11-from-rhys/wbm-e-above-mp-hull.csv"
 ).set_index("material_id")
 
-df_wrenformer["e_above_mp_hull"] = df_hull.e_above_mp_hull
-assert df_wrenformer.e_above_mp_hull.isna().sum() == 0
+df_wrenformer["e_above_hull_mp"] = df_hull.e_above_hull_mp
+assert df_wrenformer.e_above_hull_mp.isna().sum() == 0
 
 target_col = "e_form_per_atom"
 # target_col = "e_form_target"
@@ -54,7 +54,7 @@
 
     rolling_mae_vs_hull_dist(
         e_above_hull_pred=df.e_above_hull_pred,
-        e_above_hull_true=df.e_above_mp_hull,
+        e_above_hull_true=df.e_above_hull_mp,
         ax=ax,
         label=title,
         marker=marker,

mb_discovery/plots.py

@@ -31,7 +31,7 @@
     e_form="Formation energy (eV/atom)",
     e_above_hull="Energy above convex hull (eV/atom)",
     e_above_hull_pred="Predicted energy above convex hull (eV/atom)",
-    e_above_mp_hull="Energy above MP convex hull (eV/atom)",
+    e_above_hull_mp="Energy above MP convex hull (eV/atom)",
     e_above_hull_error="Error in energy above convex hull (eV/atom)",
 )
 model_labels = dict(

mb_discovery/slurm.py

@@ -28,9 +28,10 @@ def slurm_submit_python(
     array: str = "",
     pre_cmd: str = "",
 ) -> None:
-    """Slurm submit a python script using sbatch --wrap 'python path/to/file.py' by
-    calling this function in the script and invoking the script with
-    `python path/to/file.py slurm-submit`.
+    """Slurm submits a python script using `sbatch --wrap 'python path/to/file.py'`.
+    Usage: Call this function at the top of the script (before doing any real work) and
+    then submit a job with `python path/to/file.py slurm-submit`. The slurm job will run
+    the whole script.
 
     Args:
         job_name (str): Slurm job name.

models/m3gnet/join_m3gnet_relax_results.py

@@ -78,7 +78,7 @@
     f"{ROOT}/data/2022-06-11-from-rhys/wbm-e-above-mp-hull.csv"
 ).set_index("material_id")
 
-df_m3gnet["e_above_mp_hull"] = df_hull.e_above_mp_hull
+df_m3gnet["e_above_hull_mp"] = df_hull.e_above_hull_mp
 
 
 # %%

tests/test_plots.py

@@ -25,7 +25,7 @@
         f"{DATA_DIR}/{model_name.lower()}-mp-initial-structures.csv", nrows=100
     ).set_index("material_id")
 
-    df["e_above_mp_hull"] = df_hull.e_above_mp_hull
+    df["e_above_hull_mp"] = df_hull.e_above_hull_mp
 
     model_preds = df.filter(like=r"_pred").mean(axis=1)
 
@@ -62,7 +62,7 @@ def test_precision_recall_vs_calc_count(
 
         ax = precision_recall_vs_calc_count(
             e_above_hull_error=df.e_above_hull_pred,
-            e_above_hull_true=df.e_above_mp_hull,
+            e_above_hull_true=df.e_above_hull_mp,
             color=color,
             label=model_name,
             intersect_lines=intersect_lines,
@@ -96,7 +96,7 @@ def test_precision_recall_vs_calc_count_raises(
     with pytest.raises(expected_exc, match=match_pat):
         precision_recall_vs_calc_count(
             e_above_hull_error=test_dfs["Wren"].e_above_hull_pred,
-            e_above_hull_true=test_dfs["Wren"].e_above_mp_hull,
+            e_above_hull_true=test_dfs["Wren"].e_above_hull_mp,
             **kwargs,
         )
 
@@ -114,7 +114,7 @@ def test_rolling_mae_vs_hull_dist(
     ):
         ax = rolling_mae_vs_hull_dist(
             e_above_hull_pred=df.e_above_hull_pred,
-            e_above_hull_true=df.e_above_mp_hull,
+            e_above_hull_true=df.e_above_hull_mp,
             color=color,
             label=model_name,
             ax=ax,
@@ -150,7 +150,7 @@ def test_hist_classified_stable_as_func_of_hull_dist(
 
     ax = hist_classified_stable_as_func_of_hull_dist(
         e_above_hull_pred=df.e_above_hull_pred,
-        e_above_hull_true=df.e_above_mp_hull,
+        e_above_hull_true=df.e_above_hull_mp,
         ax=ax,
         stability_threshold=stability_threshold,
         stability_crit=stability_crit,